These two caskets, relic bowls, with Kharoṣṭhī inscriptions in North-western Prakrit (Gandhārī) are in the collection of Professor Samuel Eilenberg. The larger casket (Plates I–IV) of dark grey steatite is in height 10.8 cm. with seven inscribed lines, the smaller casket (Plate V) of lighter grey steatite, 7 cm. with three lines. They are of Buddhist origin and similar to the Bajaur casket published by N. G. Majumdar in Epigraphia Indica, XXIV, 1937, 1–8, and commented upon by Sten Konow in Epigraphia Indica, XXVII, 1940, 52–8.

Rajendralal Mitra was the first to edit the Sanskrit text of the Aṣṭasāhasrikā in 1888 in the Bibliotheca Indica No. 110, on the basis of six manuscripts, some of them going back to the llth century, In 1932–5 Unrai Wogihara reprinted the text in his edition of Haribhadra's Abhisamayālarikārālokā. On many occasions he, often tacitly, corrected the text from the manuscript readings, especially from ga, or from the commentary. Rather curiously, after chapter xx, p. 370, Wogihara has, for no apparent reason, omitted to indicate the variant readings of Mitra's manuscripts in his footnotes. When inserting them into my own copy of Wogihara, I was struck by Mitra's essential soundness, and by his fine instinct for Buddhist Sanskrit.

Unfortunately, not much evidence is available regarding the administrative system of Harṣa. Although we have direct sources like the travel accounts of Huien-Tsiang, the Harṣacarita of Bāṇabhaṭṭa, the Baṇskhérā and Madhubana copper-plate grants, and inscriptions of contemporary rulers, these fail to provide a comprehensive account of Harṣa's government and the political set-up of his time. We therefore supplement our knowledge with the records of the Gupta rulers. This lacuna can be filled, to some extent, if a careful study of Bāṇa's Kādambaī is made. The late V. S. Agrawala's study on Kādambaī in Hindī has a cultural bias rather than a political one. On going through the pages of Kādambaī, it appears that it is complementary to the Harṣacarita. For example, while in the Harṣacarita the picturesque account of the royal palace is mainly confined to its entrance and first block, in Kādambaī the remaining parts are dealt with at length. Similarly, while in the Harṣacarita a detailed account of all the rites which are performed on the eve of a campaign is given, in Kādambaī we get merely an indication of these activities. Thus several scenes which could not be described in the Harṣacarita owing to its limited scope, have been skilfully incorporated by the poet in Kādambaī, thus avoiding duplication. An attempt, therefore, has been made in this paper to study the political and administrative information in Kādambaī.

In 1970, while I was in search of coins and other archaeological material relating to the Śahī dynasty, I visited the town of Hund (also called Und or Ohind), the ancient Udabhāṇḍapura, the capital of the Hindu Śāhī kings. Here I made contact with Mullah Ḥabīb al-Raḥmān, a local collector of and dealer in antiquities, and from him I obtained the photograph and rubbing of a remarkable inscription of the Śāhī period. Unfortunately I was unable to see the inscription itself, since, according to the Mullah, the slab had been sent to Quetta, where it was in the hands of a private collector. Its exact provenance was uncertain, but apparently it had been discovered in the locality of Hund by workers collecting stones for building purposes.

It is nowadays common for Indian history textbooks to treat the various “empires” that successively occupied the stage of Indian history, with their respective “administrations”, as so many successive repetitions with merely different names for offices and institutions that in substance remained the same: namely, the King, the Ministers, the Provinces, the Governors, the Taxes, Landgrants, and so on. But D. D. Kosambi, in his Introduction to the study of Indian history (Bombay, 1975), rightly observed that this repetitive succession cannot be assumed, and that each regime, when subjected to critical study, displays distinct elements that call for its analysis in the context of “relations of production” (as he put it) existing at that time.

The aim of this article is to describe and assess those political changes occurring in the Godavari region of coastal Andhra (i.e. the districts of Srīkākūlam, Rajamundry, and Elluru) as a result of Mughal conquest in the later 17th century. Previous to that event, the Muslim kingdom of Golconda with its capital at Hyderabad had ruled this predominantly Telugu-speaking region. But imperial occupation did not carry with it increased political and administrative consolidation. On the contrary, extension of the boundaries of the empire brought about a perceptible decline in Muslim state power as exercised from Hyderabad. At the same time Telugu warriors who had served the king of Golconda as troop commanders and local administrative intermediaries became self-sufficient rājas.

A field of study which has hitherto been neglected in Sri Lanka is that connected with the technical aspects of arts and crafts. There is a large volume of information, mainly confined to ola manuscripts, having as their subject matter architecture, sculpture, painting, and other minor arts and crafts. In the Indian sub-continent such works dealing with the techniques of art and architecture have been studied exhaustively wherever possible and the results published in printed form. Unfortunately for us, scholars in Sri Lanka have rarely been inclined to pursue studies connected with this aspect, with the result that all relevant data have remained confined within the circles of artisans and craftsmen practising their traditional systems.